Table 4.
Summary of the characteristics of the included studies along with main results.
| Study characteristics | Participant characteristics | Interventions | Results |
|---|---|---|---|
| Arndt et al.16 RCT comparing early versus delayed initiation of passive ROM followed by formal physiotherapy Conducted in France | 92 patients (mean age = 55.3 years/37% male) Main inclusion criteria: (a) Nonretracted, isolated tear of supraspinatus repaired arthroscopically | 100 patients randomized and 92 patients followed-up (1) n = 49; early ROM, commencing day 2 post-operatively, including passive ROM, CPM without ROM limitation and daily pendular exercises (2) n = 43; maintenance of sling immobilization for 6 weeks before commencement of formal physiotherapy but still undertook daily pendular exercises | Main outcomes assessed using Constant score at 12 months: Statistically significant difference of 7.9 points (p = 0.045) in favour of early group. This difference is not regarded as clinically important No statistically significant differences between groups in terms of re-tear rate (11/49 versus 7/43; p = 0.5) |
| Cuff & Pupello17 RCT comparing early versus delayed initiation of passive ROM followed by formal physiotherapy Conducted in USA | 68 patients (mean age = 63.2 years/58% male) Main inclusion criteria: (a) Isolated full-thickness tear of supraspinatus repaired arthroscopically | (1) n = 33; early ROM, commencing day 2 post-operatively, including passive elevation and external rotation directed by a PT × 3/week and supplemented by patient directed pendular exercises between formal sessions (2) n = 35; maintenance of shoulder immobilizer for 6 weeks before commencement of formal physiotherapy but still undertook daily pendular exercises | Main outcomes assessed using American Shoulder & Elbow score at 12 months: No statistically significant differences between groups including re-tear rate (5/33 versus 3/35; p > 0.05) |
| Duzgun et al.18 RCT comparing an accelerated rehabilitation programme versus a delayed programme Conducted in Turkey | 29 patients (mean age = 56.3 years/10% male) Main inclusion criteria: (a) Rotator cuff rupture repaired arthroscopically | (1) n = 13; early passive ROM, commencing day 7 post-operatively, followed by active ROM commencing day 21 and resistance from day 28. (2) n = 16; delayed programme with active ROM commencing day 42 post-operatively | Main outcomes assessed using: Disabilities of the Arm, Shoulder & Hand at 8 weeks, 12 weeks, 16 weeks and 24 weeks: Statistical (p < 0.05) and clinically (>10 points) significant difference in favour of the accelerated group at 8 weeks, 12 weeks and 16 weeks but no significant difference by 24 weeks |
| Hayes et al.25 RCT comparing a standardized home exercise programme plus individualized treatment versus a standardized home exercise programme alone Conducted in Australia | 58 patients (mean age = 60.2 years/71% male) Main inclusion criteria: (a) Diagnosis of rotator cuff rupture, of any size repaired surgically | (1) n = 26; sling immobilization for 1 day post-operatively followed by encouragement to commence light functional activity and pendular exercises for further 7 days. Active-assisted ROM from day 8 onwards and active and resisted exercise commenced from day 42 onwards. Supplemented by individualized physiotherapy from second week post-operatively including exercise, MT, ET at the discretion of the treating physiotherapist (2) n = 32; standardized home exercise programme alone | Main outcomes assessed using Shoulder service questionnaire (SSQ) at 6, 12 and 24 weeks: No statistically significant differences between groups across all time points except physical symptoms, lifestyle and overall shoulder status domains of SSQ at 24 weeks in favour of home exercise plus individualized treatment group. Clinical importance of this difference is unclear |
| Keener et al.19 RCT comparing early passive ROM versus delayed ROM with sling immobilization for 6 weeks Conducted in USA | 124 patients (mean age = 55.3 years/59% male) Main inclusion criteria: (a) <65 years of age (b) Diagnosis of full thickness rotator cuff tear <30 mm repaired arthroscopically | 1. n = 65; pendular exercises immediately post-operatively and therapist supervised passive ROM from 7 days post-operatively. Active ROM initiated from day 42 onwards 2. n = 59; shoulder immobilized for 6 weeks post-operatively before commencement of therapist supervised passive ROM | Main outcomes assessed using American Shoulder & Elbow score at 6 months, 12 months and 24 months: No statistically significant differences between groups including re-tear rate (6/63 versus 3/53; p = 0.46) |
| Kim et al.20 RCT comparing early passive ROM versus delayed ROM with brace immobilization for 5 weeks Conducted in South Korea | 105 patients (mean age = 60 years/42% male Main inclusion criteria: (a) Diagnosis of small to medium-sized full-thickness rotator cuff tears repaired arthroscopically | (1) n = 56; abduction brace for up to 35 days post-operatively supplemented by passive ROM 3 to 4 times per day during this period (2) n = 49; abduction brace only with no passive motion during this period | Main outcomes assessed using American Shoulder & Elbow score at 6 months and 12 months: No statistically significant differences between groups including re-tear rate (7/56 versus 9/49; p = 0.43) |
| Klintberg et al.8 RCT comparing early loading versus delayed loading Conducted in Sweden | 14 patients (mean age = 55 years/64% male) Main inclusion criteria: (a) Diagnosis of full-thickness tear repaired surgically | (1) n = 7; low-level active ROM × 3/day from day 2 post-operatively supplemented by passive ROM directed by the physiotherapist. Load was progressed from day 28 post-operatively when sling immobilization was ceased. (2) n = 7; 6 weeks of sling immobilization supplemented by passive ROM | Main outcomes assessed using Constant score at 6, 12 and 24 months: Between group difference inadequately reported; reported as no difference in adverse effects but statistical significance unclear |
| Koh et al.21 RCT comparing immobilization for four versus eight weeks Conducted in South Korea | 100 patients (mean age 59.9 years/50% male) (a) Diagnosis of full-thickness tear, 2 cm to 4 cm in size, repaired arthroscopically | (1) n = 47; 4 weeks of immobilization without passive ROM (2) n = 53; 8 weeks of immobilization without passive ROM | Main outcomes assessed using Constant score and American Shoulder and Elbow Surgeons score at 6 months and 24 months: No statistically significant differences between groups including re-tear rate (5/40 versus 4/48; p = 0.73) |
| Lastayo et al.22 RCT comparing continuous passive motion versus manual passive ROM exercises Conducted in USA | 31 patients (mean age 63.3 years/44% male) (a) Rotator cuff tear repaired surgically | (1) n = 17; home continuous passive motion for 4 hours per day after discharge from hospital for 4 weeks, supplemented by daily pendular exercises (2) n = 15; manual passive ROM exercises three times per day performed by carer or similar for 4 weeks supplemented by daily pendular exercises | Main outcomes assessed using Shoulder Pain and Disability Index at unclear time point: No statistically significant (p > 0.05) differences between groups |
| Lee et al.23 RCT comparing aggressive versus limited passive exercises Conducted in South Korea | 64 shoulders (mean age 54.9 years/64% male) (a) Diagnosis of medium- or large-sized full-thickness rotator cuff tear repaired arthroscopically | (1) n = 30; immediate passive ROM × 2/day without limit on ROM supplemented by daily pendular exercises with shoulder brace maintained in situ for 6 weeks (2) n = 34; continuous passive movement limited to 90° × 2/day and passive ROM with shoulder brace maintained in situ for 6 weeks | Main outcomes assessed using University of California Los Angeles shoulder rating scale at 3 months and 6 months: Statistically significant (p < 0.01) difference in favour of aggressive exercise at 3 months but unknown if difference of 2.9 points is clinically significant. No statistically significant difference by 6 months (p = 0.16). No statistically significant difference between groups in terms of re-tear rate (7/30 versus 3/34; p = 0.11) |
| Raab et al.24 RCT comparing physiotherapy versus physiotherapy with continuous passive motion Conducted in USA | 26 patients (mean age 55.8 years/69% male) (a) Rotator cuff tear repaired surgically | (1) n = 12; physiotherapy (no further description) (2) n = 14; physiotherapy with continuous passive movement commencing in the recovery room, progressed within pain-free limits, and continuing for 8 hours/day for 3 weeks limited to 90° × 2/day and passive ROM with shoulder brace maintained in situ for 6 weeks | Main outcomes assessed using an author generated patient-reported shoulder score at 3 months: No statistically significant difference between groups (p = not reported) |
| Roddey et al.33 RCT comparing two approaches to home exercise instruction Conducted in USA | 108 patients (mean age 58 years/64% male) (a) Diagnosis of full-thickness tear repaired arthroscopically | (1) n = 54; videotape based home exercise instruction while sling remained in situ for 6 weeks. Passive exercise for 4 weeks to 6 weeks, followed by active exercise between 6 weeks and 12 weeks and then strengthening exercises >3 months (2) n = 54; personal PT instruction while sling remained in situ for 6 weeks. Principles of exercise progression as group 1 | Main outcomes assessed using Shoulder Pain & Disability Index at 3 months, 6 months and 12 months: No statistically significant difference between groups (p = 0.17, 0.40, 0.99 respectively) |
RCT, randomized controlled trial; ROM, range of motion; PT, physiotherapist/physical therapist; MT, manual therapy; ET, electrotherapy including heat and ice.